def fit_basic_cd(modelspecs,
est,
maxiter=1000,
ftol=1e-8,
IsReload=False,
**context):
'''
A basic fit that optimizes every input modelspec. Use coordinate
descent for fitting and nmse_shrink for cost function
'''
if not IsReload:
fit_kwargs = {'options': {'ftol': ftol, 'maxiter': maxiter}}
if type(est) is list:
# jackknife!
modelspecs_out = []
njacks = len(modelspecs)
i = 0
for m, d in zip(modelspecs, est):
i += 1
log.info("Fitting JK {}/{}".format(i, njacks))
metric = lambda d: metrics.nmse_shrink(d, 'pred', 'resp')
modelspecs_out += nems.analysis.api.fit_basic(
d,
m,
fit_kwargs=fit_kwargs,
metric=metric,
fitter=coordinate_descent)
modelspecs = modelspecs_out
else:
# standard single shot
# print('Fitting fit_basic')
# print(fit_kwargs)
metric = lambda est: metrics.nmse_shrink(est, 'pred', 'resp')
modelspecs = [
nems.analysis.api.fit_basic(est,
modelspec,
fit_kwargs=fit_kwargs,
metric=metric,
fitter=coordinate_descent)[0]
for modelspec in modelspecs
]
return {'modelspecs': modelspecs}
def fit_cd_nfold_shrinkage(modelspecs, est, IsReload=False, **context):
''' fitting n fold, one from each entry in est, use mse_shrink for
cost function'''
if not IsReload:
metric = lambda d: metrics.nmse_shrink(d, 'pred', 'resp')
modelspecs = nems.analysis.api.fit_nfold(est,
modelspecs,
metric=metric,
fitter=coordinate_descent)
return {'modelspecs': modelspecs}
def fit_basic(modelspecs,
est,
max_iter=1000,
tolerance=1e-7,
shrinkage=0,
IsReload=False,
**context):
''' A basic fit that optimizes every input modelspec. '''
if not IsReload:
if shrinkage:
metric = lambda d: metrics.nmse_shrink(d, 'pred', 'resp')
else:
metric = lambda d: metrics.nmse(d, 'pred', 'resp')
fit_kwargs = {'tolerance': tolerance, 'max_iter': max_iter}
if type(est) is list:
# jackknife!
modelspecs_out = []
njacks = len(modelspecs)
i = 0
for m, d in zip(modelspecs, est):
i += 1
log.info("Fitting JK {}/{}".format(i, njacks))
modelspecs_out += nems.analysis.api.fit_basic(
d,
m,
fit_kwargs=fit_kwargs,
metric=metric,
fitter=scipy_minimize)
modelspecs = modelspecs_out
else:
# standard single shot
modelspecs = [
nems.analysis.api.fit_basic(est,
modelspec,
fit_kwargs=fit_kwargs,
metric=metric,
fitter=scipy_minimize)[0]
for modelspec in modelspecs
]
return {'modelspecs': modelspecs}
def fit_nfold_shrinkage(modelspecs,
est,
ftol=1e-7,
maxiter=1000,
IsReload=False,
**context):
''' fitting n fold, one from each entry in est, use mse_shrink for
cost function'''
if not IsReload:
metric = lambda d: metrics.nmse_shrink(d, 'pred', 'resp')
fit_kwargs = {
'options': {
'tolerance': tolerance,
'max_iter': max_iter
}
}
modelspecs = nems.analysis.api.fit_nfold(est,
modelspecs,
metric=metric,
fitter=scipy_minimize,
fit_kwargs=fit_kwargs)
return {'modelspecs': modelspecs}
def fit_basic_shr_init(modelspecs, est, IsReload=False, **context):
'''
Initialize modelspecs in a way that avoids getting stuck in
local minima.
written/optimized to work for (dlog)-wc-(stp)-fir-(dexp) architectures
optional modules in (parens)
'''
# only run if fitting
if not IsReload:
metric = lambda data: metrics.nmse_shrink(data, 'pred', 'resp')
modelspecs = [
init.prefit_LN(est,
modelspecs[0],
analysis_function=nems.analysis.api.fit_basic,
fitter=scipy_minimize,
metric=metric,
tolerance=10**-5,
max_iter=700)
]
return {'modelspecs': modelspecs}